Manual de Chess King 2015 | Page 114

PART 1 • Opening choices at a glance Let's step through the tree, following those moves, and see what we get when we reach that position: • You'll recall that after 11.Bd2, Houdini 2 recommended the move 11... Ngf6. But that's the second-favorite move on the list right? • Look closer. While it's played less often than 11...e6, it's performed a bit better statistically than the Knight move. Meanwhile, a chess engine evaluates the two moves as being equal – a bit better than a tenth of a pawn in White's favor, close enough to zero to call the position an even game. • The tree also shows three other moves, and while 11...Ndf6 scored a 50% success percentage, we need to note that it was played just once, meaning the game was a draw, and that a chess engine evaluated the position as being nearly a half-pawn in White's favor (a nice edge for White). So I think we can safely reject that idea. That brings us to another important point: when looking at a move's statistics in a chess tree, take any “success” numbers with a big dose of salt if that move has been played very few times. In the case of the aforementioned 11...Ndf6, it scores well for White according to a chess engine, but the game was still a draw – it's entirely possible that Black squeaked by with a draw because of a White mistake later in the game. It's just another example of why statistical data should be used as a guide, a signpost, but it's never the last word – you should always think for yourself! 114 chessking.com